Closed GoogleCodeExporter closed 9 years ago
I can reproduce and confirm this issue. Though, I'm undecided if this is
something that needs fixing. Since this is a crypto library, it was only ever
meant to operate on whole bytes.
Original comment by Jeff.Mott.OR
on 5 Oct 2011 at 6:03
I agree with Jeff especially since it is unclear what would be expected. 12F
could easily be expected to be [0x12, 0xF0] instead of [0x1, 0x2F] depending on
whether you expect the padding 0 to be first or last. An even number of digits
should be considered invalid input to this function and the only reasonable
change would be to throw but since this is a low level routine even that test
isn't justified.
Original comment by Chuck...@gmail.com
on 26 Oct 2011 at 8:05
Original comment by Jeff.Mott.OR
on 2 Nov 2011 at 11:44
Original issue reported on code.google.com by
sam...@erdtman.se
on 4 Oct 2011 at 12:58